N2048P ECC error

Question

We just got 10 N2048P and we were going to try one of them in the current enviroment before deploying all of them, unfortuntaly we got those errors in the log:

Apr 28 09:58:08 SW-SPES-06-1 DRIVER[bcmDPC]: broad_hpc_drv.c(4428) 366 %% Unit: 0 #012 CDC RX FIFO entry 44 double-bit ECC error
Apr 28 09:58:08 SW-SPES-06-1 DRIVER[bcmDPC]: broad_hpc_drv.c(4428) 367 %% Unit: 0 #012 CDC RX FIFO table 0 entry 44 ECC error
Apr 28 09:59:50 SW-SPES-06-1 DRIVER[bcmDPC]: broad_hpc_drv.c(4428) 369 %% Unit: 0 #012 CDC RX FIFO entry 28 double-bit ECC error

our network monitor says that some of the devices attached are losing pings every now and then (even if we can't see any packet loss with a manual ping) and it gets this error from the switch via SNMP:

CRIT - [Unit: 1 Slot: 0 Port: 22 Gigabit - Level] (up) MAC: f8:b1:56:62:b2:fa, 10MBit/s, in: 57.19B/s, in-errors: 2.78%(!!) = 0.1, out: 4.68kB/s

Note that the device attached to the port 22 is NOT one that gave us warnings about pings

I've seen here that someone got the same problem on the same switch and it could be related to a memory problem, I upgraded the firmware from 6.1.0.6 to 6.2.0.5 but no luck, I decided to try another switch but I got the same problem (i copied the configuration), so it seems that the problem is not really related to the switch.

Before the N2048P we were using a small Netgear Prosafe FS116P without any apparent problem (it is unmanaged tho), any though on what it could be or on how can it debug the problem?

DELL-Josh Cr · Answer

Hi,

If you copied the config to another switch and the other switch worked before it isn’t an issue with the memory in the switch. How many devices are currently connected? If you disconnect other devices does the error continue? How often is it occurring? If you clear the config do the errors still happen?

giovanni.lnl · Answer

Hi,

thanks for your reply and sorry for my late one!

Unfortunatly we needed to have a working network for a deadline so we just replugged everything in the Netgear POE switch.

Now i'm going to try to replicate the problem, i got myself some spare devices was attached to the switch and that, during the previous tests, had some ping problems, they are embeded linux system (www.expertdaq.com/.../ex-9486-l) wich runs some kind of control software.

What was attached to the switch during the tests was 4 of them plus 2 poe access points and 2 other non-poe systems, we noticed the problems only when most of them were attached, if we unplug a couple of devices the erorrs rate was a lot lower and no ping test was failing.

We ruled out a poe problem because i tried to attach some of the poe devices to the Netgear switch which was attached to the Dell one, same error and same problem.

The errors occur quite often, i'd say one every 1 or 2 second in the worst case, with less devices attached it was one every 3 or 4 minutes, I tried to clear the configuration and test it again but nothing changed.

Could you please tell me what the error is about? the "ECC" part made me thought about some memory problem but i doubt is about that.

Thanks

DELL-Josh Cr · Answer

If it is doing it even with a different switch then it probably is not a hardware issue and is something else in your setup that is causing it. ECC is memory error correction, so it may not be causing an actual problem if it is able to correct the error. It could be something passing from the embedded linux devices.

giovanni.lnl · Answer

I fear the problem is something related to the network traffic generated from the software running on the embedded devices, i tried 4 of them on one of the switches that gave us the errors but without any problem.

unfortunatly at the moment i can't install that software on them so i'll have to wait to be able to do that and see if i can replicate the problem,

Networking General

N2048P ECC error

Was this post helpful?